Evaluation Model of Disaster Rescue Effectiveness based on AHP

This study aims to address challenges faced by visually impaired individuals in daily travel, such as inability to perceive traffic lights, navigate obstacles, and locate tactile pavings. The goal is to develop a specialized travel assistance algorithm for visually impaired individuals. After thorough research on existing technologies including GPS, IrDA, AI-assisted technology, mobile applications, and object recognition, and analyzing their limitations, we proposed a blind navigation algorithm based on the PP-YOLO model. This algorithm combines traffic sign recognition, object recognition, and tactile paving recognition to achieve multi-task image recognition. Through dataset preparation, data annotation, and parameter adjustment, the algorithm was implemented on portable devices like mobile phones and wearable devices. Laboratory verification and field testing confirmed that the algorithm provides accurate navigation guidance and ensures safety for visually impaired individuals during travel. With a frame rate of 8fps, the algorithm achieved a recognition rate of 96.3% within an 8 m range on portable devices; in low-light environments, the object recognition accuracy within the same range reached 94%. Even at complex intersections with a width of 30 m, the recognition rate was 85%. This research outcome presents an innovative technical solution for travel assistance technology for visually impaired individuals and offers new research ideas for similar studies. The PP-YOLO model was chosen for its superior speed and accuracy in object detection tasks, crucial for real-time navigation. The model was streamlined for use on portable devices through lightweight design and hardware acceleration features such as NPU and GPU. The challenge with this approach involves some risk of recognition errors, similar to those in autonomous driving technology. However, considering the slower walking speed of visually impaired individuals, the consequences of these errors are mitigated. Thus, the project provides an additional mobility option for visually impaired individuals, providing more freedom than without it.